Neuroscientists are faced with a torrent of experimental data about the structure and function of the brain in health and disease. Much of this information pertains to the cerebral cortex and cerebellar cortex, the dominant structures of the human brain. These are challenging to study because they are highly convoluted sheet-like structures and are highly variable from one individual to the next. We propose a set of major enhancements to our existing set of neuroinformatics software tools for visualization, analysis, and data mining that facilitate studies of cerebral and cerebellar cortex. Key components of this resource include (i) surface- based atlases in humans and nonhuman primates that capture the shape of cortical convolutions and represent population averages as well as individual variability; (ii) a database of neuroimaging data (SumsDB) coupled to an online visualization tool (WebCaret); and (iii) a brain-mapping software application (Caret) that provides numerous analysis and visualization capabilities. Our atlasing efforts will concentrate on human cerebral cortex, which contains a complex but poorly understood mosaic of distinct areas (100-200 in total). We will generate a 'library' of cortical areas in SumsDB representing many different published maps of cortical areas, identified using a variety of methods and all registered to a common atlas framework. This library will serve as a valuable and easily accessible reference collection for neuroscientists studying functional localization. We will also generate libraries of published experimental data related to cortical function and structure in normal individuals and in a variety of disease conditions. These libraries will facilitate cross-study comparisons that reveal consistent functional characteristics of each cortical area and functional differences between nearby areas. We will compare the functional organization of human cerebral cortex to that in the chimpanzee and macaque monkey. This will provide insights as to which cortical areas have a common evolutionary origin in each species, which areas have expanded in size in the human lineage, and whether there are cortical areas present in humans that are altogether absent in other primate species. Finally, we will implement several major enhancements in Caret software. This will include methods for more efficient and sensitive characterization of normal structure and function and of cortical abnormalities in disease conditions. Successful execution of these aims will provide neuroscientists with a powerful and flexible set of tools for analyzing, visualizing, sharing, and accessing a broad range of experimental data, thereby accelerating efforts to better understand, diagnose, and treat many brain disorders.